Objective 1: To demonstrate effective integrated approaches to Brassica pest management
Due to the unavoidable delay to the start of project activities, the workshops and farmer extension sessions were postponed; they will now be held in early September, 2006. Despite the delay, several stakeholder and technical meetings were held in November 2005; these established an introduction to the project and preliminary planning sessions were conducted (additional technical meetings were held in June 2006 ; Objective 3).

Seminars and discussion groups were held to introduce stakeholders to the project's approach and goals at:
SPC, Suva (Fiji). The meeting was attended by approximately 20 stakeholders including farmers, SPC and MASLR extension officers and officers from the Fiji Ministry of Health. In an additional meeting, technical aspects of the project were discussed by MASLR, SPC, University of Queensland and QDPI&F project team members.

Nu'u crop research station (Samoa). The meeting was attended by approximately 15 stakeholders including farmers, MAF, SPC, University of Queensland and QDPI&F technical and extension staff. In a supplementary technical meeting, project team members (MAF, SPC, UQ and QDPI&F) discussed issues related to the project for the coming 2006 season.

The participatory trials to demonstrate the essential elements of effective Brassica IPM trials have not yet been set up but they will be conducted to coincide with the planned workshops in early September in both Fiji and Samoa.

Objective 2: The introduction of a FFS approach for improved sustainable production systems in Brassica crops and effective communication of project outputs
A possible master trainer for the training of local extension officer personnel in FFS techniques has been identified, training session are expected to begin in the current crop season. Appropriate community/ farmer groups have been identified for the basis of FFS in Fiji and Samoa. It is anticipated that training will have begun and be ongoing by the time of the September workshops; the first field testing of the toolkit will be undertaken around the workshops. Team members in Fiji and Samoa regularly liaise with DSAP team members. A key member of the IPPSI project team is now a PhD student at the University of Queensland, this close link between personnel between both projects will ensure effective bilateral information exchange and cooperation.

Objective 3: To determine the role of major natural enemies in the management of Brassica crop pests
Regional surveys (on farms and research stations) to determine the key insect pests of Brassica crops and their associated natural enemies began in June 2006 in both Fiji and Samoa. Continuous data collection for the next two years (June 2006- May 2008) will build up this inventory. Local scientists have received training relating to survey techniques and insect identification.

Preparations for experiments to determine the impact of relevant natural enemies on populations of diamondback moth are underway in both countries. The experiments will be performed later this season and will provide valuable preliminary information regarding the structure and function of the endemic natural enemy complexes in the Brassica agro-ecosystems of both countries.

Preliminary data indicate that some important natural enemies of the diamondback moth are present in both countries. Arrangements have been made to receive a population of a heat tolerant biotype of Diadegma insulare from Florida into the quarantine facility at the University of Queensland. Experiments have investigated the relative foraging efficiencies and survival of adult and immature D. semiclausum at a range of temperatures. Preliminary results indicate that D. semiclausum forages effectively at temperatures up to 25C, however at temperatures of 25C and above successful development to adult stages is severely disrupted. Suitable protocols for comparing the performance and survival of D. semiclausum and D. insulare at a range of temperatures in the laboratory have been devised.

Objective 4: To develop a refined local IPM strategy using selective plant protection products
Selective plant protection products have been sourced and transported to Fiji and Samoa for testing. Products will initially be tested in the laboratory and then field tests will be conducted accordingly. In Australia, laboratory tests on the relative efficacy of a Neem formulation against P. xylostella and C. pavonana are currently underway. These preliminary experiments will be completed by the time of the workshops in September and results will be discussed with team members.

Current action thresholds for intervention are based on the assessment "standard" insect measurements. These relative values are based on the total foliage consumption by first to final stage larvae of each species. Preliminary experiments measured the relative foliage consumption of P. xylostella and C. pavonana in the laboratory. Individual C. pavonana larvae consumed approximately 13-times more cabbage foliage than individual P. xylostella larvae. Based on the standard insect assessment, one C. pavonana larva represents 2.6 standard insects.

Objective 1: To demonstrate effective integrated approaches to Brassica pest management
The postponed workshops and farmer extension sessions were conducted in September 2006. They were attended by more than 30 participants in Samoa and more than 150 participants in Fiji. Participatory trials to demonstrate the essential elements of effective Brassica IPM were set up in both countries in June 2007 and followed on from the FFS workshops (see Objective 2 below). Further participatory trials are underway and will be completed in September/ October 2007.

Objective 2: The introduction of a FFS approach for improved sustainable production systems in Brassica crops and effective communication of project outputs
A FFS master trainer from the FAO IPM programme in the Philippines ran two-week long training activities in Samoa (May 21-31, 2007) and Fiji (June 4- 15, 2007) to acquaint local extension officers with FFS techniques. In Samoa 24 participants took part in the training and in Fiji 22 participants were involved. During the training activities the participants formed groups to conduct a pilot FFS on a research station in each country following the intensive two-week long training activities. To ensure that the pilot FFS activities were adequately supported in their early stages, the master trainer visited each FFS group approximately two weeks after the completion of the training.

Objective 3: To determine the role of major natural enemies in the management of Brassica crop pests
Regional surveys to determine the key insect pests of Brassica crops and their associated natural enemies began in June/ July 2006 in both Fiji and Samoa. Data on pest abundance was collected weekly at experimental sites in both countries. In Samoa diamondback moth (DBM) abundance was high in July and September 2006 but declined to low levels by mid February 2007 and the pest was not detected in the field after April 2007. Large cabbage moth (LCM) populations were high in August and September 2006 but declined through December, January and February before reaching high levels again in April through July 2007. Pest surveys in Fiji have been more extensive (conducted weekly at three research stations) and have recorded more pests than the surveys in Samoa. In Fiji in addition to DBM and LCM, Spodoptera littura, Trichoplusia ni, Helicoverpa armigera and aphids have been recorded on Brassica crops. At Koronivia research station DBM populations were high from July through December 2006 but remained very low through to the end of June 2007. At Legalega research station the DBM population reached a maximum in October 2006 but the pest was only recorded infrequently at this site from November 2006 to June 2007. At Sigatoka research station DBM populations reached a maximum in December 2006 and the pest was only recorded at very low densities at the site between December 2006 and June 2007. LCM was recorded at Koronivia and Leglega research stations throughout the surveys but it was not recorded at Sigatoka after February 2007.

Collections of DBM in Fiji have confirmed that the larval parasitoid Cotesia vestalis (=plutellae) Haliday and the larval pupal parasitoid Oomyzus sokolowskii Kurdjumov are established at all three sites. Collections of DBM in Samoa have confirmed the presence of C. vestalis and O. sokolowskii in the country. Both C. vestalis and O. sokolowskii are effective at higher temperatures than D. semiclausum and have the potential to be useful biological control agents in both countries; their impact on DBM populations will be measured. The recovery of these parasitoids from DBM in both Fiji and Samoa indicates that releases of exotic parasitoids are unlikely to be necessary in either country.
Only limited collections of LCM have been undertaken in Fiji and no egg or larval parasitoids have been recovered. Limited collection and rearing of LCM larvae has failed to detect any larval parasitoids of this pest in Samoa but Trichogramma chilonis has been discovered parasitising egg masses of LCM.

Natural enemy exclusion experiments in Fiji have revealed the importance of predators (and possibly abiotic factors) as causes of mortality in experimental populations of DBM.

Objective 4: To develop a refined local IPM strategy using selective plant protection products
During the FFS training courses, preliminary IPM management programmes were devised for Fiji and Samoa. These programmes, which are based on the threshold directed application of Bt in response to the combined pest complex, were tested by members of the FFS and compared with control and current farmer practice. The experiments will be completed by late August and reported on in the next annual report.

Objective 1: To demonstrate effective integrated approaches to Brassica pest management
A trial conducted at Sigatoka research station tested a preliminary IPM strategy against current farmer practice and control (no intervention) treatments. The experiment showed that IPM effectively managed the pest complex (diamondback moth (DBM) and large cabbage moth (LCM)), promoted natural enemy activity and resulted in crop yields which were  yields achieved by farmer practice. A similar experiment was conducted at Nu'u crop research station in Samoa March- May 2008 but the data is yet to be received. A refined version of the preliminary IPM strategy will be tested in both Fiji and Samoa in the 2008 growing season.

Objective 2: The introduction of a FFS approach for improved sustainable production systems in Brassica crops and effective communication of project outputs
Following the FFS train the trainer activities (Samoa: May 21-31, 2007; Fiji: June 4- 15, 2007) reported on previously the master trainer revisited each FFS group shortly after its inception to help deal with any problems and provide addition support during the early stages of activities. The FFS groups in both countries were not sustained following the departure of the master trainer. The FFS train the trainer activities for 2008 have been postponed until July and August as inclement weather in Fiji has delayed the preparation of field training plots. The sustainability of FFS activities following the departure of overseas project team members will be specifically addressed in the forthcoming training activities. Project outputs were communicated by oral and poster presentations at the
Australia and New Zealand Biocontrol Conference, Sydney 10-14 February 2008. In 2007 a documentary explaining the aims and objectives of the project was broadcast on national television networks throughout the Pacific Island countries and territories. In December 2007 the project leader traveled to Solomon Islands and worked closely with entomologists involved in ACIAR project HORT/2003/047 to develop approaches to survey Brassica crops for pest insects and their natural enemies.

Objective 3: To determine the role of major natural enemies in the management of Brassica crop pests
Regional surveys have confirmed that the DBM parasitoids Cotesia vestalis and Oomyzus sokolowskii are widely established in the major Brassica crop growing areas of Viti Levu, Fiji. The surveys did not record any parasitoids of LCM. In September 2007 a large study to measure the impact of natural enemies on experimental populations of DBM was conducted at Sigatoka research station. The work showed that both C. vestalis and O. sokolowskii can be effective parasitoids of DBM in Fiji and that considerable pest mortality also occurs in cabbage crops due to the combined action of generalist predators (predominantly spiders and ants). In Samoa the establishment of C. vestalis and O. sokolowskii has been confirmed but no studies have yet measured their impact on pest populations (extremely low field populations of the pest have been recorded following its seasonal decline in the 2007 wet season). Regional surveys have recorded LCM egg masses attacked by Trichogramma chilonis at a number of sites on Upolo and preliminary studies to assess the impact of the egg parasitoid on its host indicate that it may have a significant role to play in an IPM programme.

Objective 4: To develop a refined local IPM strategy using selective plant protection products
Preliminary IPM programmes based initially on threshold directed intervention with foliar applications of Bacillus thuringiensis have been developed and tested. Initial studies indicate that DBM thresholds are appropriate but that LCM thresholds will need some revision, revised thresholds and a modified sampling plans will be tested in the 2008 field season. The efficacy of a range insecticides (Prevathon (rynaxphur), Steward (indoxacarb), Match (lufenuron), Delfin (Bacillus thuringiensis - kurstaki) and Xentari (Bacillus thuringiensis- aizawi) was tested against the insect pest complex of Brassica crops in a field experiment conducted at Sigatoka research station in the latter half of 2008. The performance of these compounds was compared to that of a synthetic pyrethroid (prophylactic application of which is currently standard practice for most farmers in both Fiji and Samoa). The selective insecticides performed better than the synthetic pyrethroid and the study indicated that both formulations of Bacillus thuringiensis are suitable for inclusion in IPM strategies against the Brassica pest complex in Fiji.

Good progress has been made towards demonstrating the effectiveness of integrated pest management (IPM) approaches to Brassica pests in both Fiji and Samoa. In 2007 a trial conducted at Sigatoka research station in Fiji tested a preliminary IPM strategy against current farmer practice and control (no intervention) treatments. The experiment showed that IPM effectively managed the pest complex (diamondback moth (DBM) and large cabbage moth (LCM)), promoted natural enemy activity and resulted in crop yields which were greater than or equal to yields achieved by farmer practice. Similar experiments were conducted at Nu'u crop research station in Samoa between August and October and November and December 2008 and from May 2009 (ongoing). These experiments confirm that the IPM strategy promotes natural enemy activity and can maintain pest numbers below threshold levels and that far fewer applications of a biological pesticide based on Bacillus thuringiensis (Bt) are required than the number of conventional insecticide applications typically applied as standard farmer practice.

The farmer field school (FFS) approach to participatory learning for sustainable Brassica production systems has been successfully re-established in both countries after initial setbacks. Following the FFS train the trainer activities in 2007, FFS groups were not sustained in either country upon the departure of the master trainer. The FFS train the trainer activities in 2008 specifically addressed the sustainability of FFS activities following the departure of overseas project team members. The approach was extremely successful and four FFS (three in Fiji and one in Samoa) were completed in 2008. Following the success of these FFS activities, extension officers trained in the project's train-the-trainer program have established FFSs with farmer groups at two villages in Fiji and at one village in Samoa. These activities are continuing.

Studies to determine the role of major natural enemies in managing Brassica pests have yielded somewhat different results in the two countries. Regional surveys have confirmed that the DBM parasitoids Cotesia vestalis and Oomyzus sokolowskii are widely established in the major Brassica crop growing areas of Viti Levu, Fiji, and Upolu, Samoa. Exclusion studies have shown that natural enemies can have a significant impact on pest populations in both countries; however, while the impact of C. vestalis and O. sokolowskii on DBM populations have been quantified in Fiji, similar experiments have not yet quantified the impact of these parasitoids in Samoa. On the other hand, field studies in Samoa have established the impact of the egg parasitoid Trichogramma chilonis on LCM and the importance of ants as predators of DBM.

An IPM approach based on threshold-directed intervention with foliar applications of Bt has been developed and tested in both countries. Initial studies indicate that DBM thresholds are appropriate but that LCM thresholds will need some revision; a revised threshold is currently being tested in Samoa. The efficacy of several insecticides (Prevathon (rynaxphur), Steward (indoxacarb), Match (lufenuron), Delfin (Bt-kurstaki) and Xentari (Bt-aizawi) was tested against the insect pest complex of Brassica crops in a field experiment conducted at Sigatoka research station in Fiji. The performance of these compounds was compared to that of Suncis (a synthetic pyrethroid that is currently applied as a prophylactic treatment by most farmers in both Fiji and Samoa) and an untreated control. The most widely used insecticides (Suncis, Steward and Match) did not perform significantly better than the untreated control, indicating that they are ineffective in the field in Fiji. Performance of both formulations of Bt was better, indicating that they are suitable for inclusion in IPM strategies against the Brassica pest complex in Fiji.

Laboratory studies have confirmed that DBM in Fiji has developed resistance to conventional insecticides currently being used by farmers. A field population of diamondback moth collected from Sigatoka displayed evidence of resistance to Match (lufenuron) and Suncis (cypermethrin) when compared to a standard susceptible population (Waite strain). However, there was no evidence for resistance to Dipel (Bt-kurstaki), which has been utilised only to a very limited extent in Fiji and which is integral to the IPM approach being developed in both Fiji and Samoa. Sub populations of the DBM collected in Sigatoka have been selected with different insecticides (Match, Suncis and Steward) and current experiments are examining these for evidence of increased resistance and cross resistance.

Studies to elucidate to host-plant preferences of LCM are under way. In laboratory studies, sexually mature mated LCM females oviposited more readily on Chinese cabbage plants (all females tested oviposited on exposed plants on the first night of exposure) than on head cabbage plants (median time to oviposition of first egg mass was 5 days) in no-choice experiments; however, there was no difference between the size of egg masses laid on the two host plants (mean egg mass size (SE) was 48 (7) on head cabbage, and 59 (17) on Chinese cabbage). Preliminary olfactometer experiments show that female LCM are attracted to the volatiles produced by head cabbage plants; Chinese cabbage plants are yet to be tested.

In further studies to evaluate the potential of T. chilonis as a biological control agent of LCM, following on from egg parasitism studies in 2007, three field experiments were conducted to assess the impact of T. chilonis on LCM eggs in the field at Nu'u crops research station in Samoa. In the first (conducted 18-22 August) the field-laid cohort suffered significant mortality, with 44% of egg masses disappearing before hatching, but none was parasitised by T. chilonis. In the second (conducted 11-15 November) the field-laid cohort (n=57) also suffered significant mortality; in this instance only 4.5% of egg masses disappeared before hatching but 66% were attacked by T. chilonis. In the third experiment (conducted 13-18 May) the field-laid cohort (n=57) again suffered significant mortality, 72% of egg masses disappearing before hatching, but none was parasitised by T. chilonis. Thus, T. chilonis can exert significant, but highly variable, levels of mortality on LCM eggs in cabbage crops in Samoa.

Further progress has been made towards demonstrating the effectiveness of integrated pest management (IPM) approaches to Brassica pests in both Fiji and Samoa. A trial conducted at Sigatoka research station in Fiji tested a preliminary IPM strategy against current farmer practice and control (no intervention) treatments. The experiment showed that IPM effectively managed the pest complex (diamondback moth (DBM) and large cabbage moth (LCM)), promoted natural enemy activity and resulted in crop yields which were greater than or equal to yields achieved by farmer practice. Several similar experiments have been conducted at Nu'u crop research station in Samoa (August-October 2008, November-December 2008, May-June 2009, August-September 2009 and April-June 2010). These experiments confirm that the IPM strategy promotes natural enemy activity, can maintain pest numbers below threshold levels and that far fewer applications of a Bacillus thuringiensis (Bt) are required than the number of conventional insecticide applications typically applied as standard farmer practice. The decision threshold is based on the integration of independently calculated DBM and LCM thresholds to determine a pest index for intervention. Studies to date indicate that DBM thresholds are appropriate but that LCM thresholds could be improved. LCM thresholds are currently being tested in Samoa and field experiments in 2010 will directly compare a range of thresholds for interventions with Bt and Prevathon.

Following the Farmer Field School (FFS) train-the-trainer activities in 2007, FFS groups were not sustained in either country upon the departure of the master trainer. The FFS train-the-trainer activities in 2008 specifically addressed the sustainability of FFS activities following the departure of overseas project team members. The approach ensured that four FFS (three in Fiji and one in Samoa) were completed in late 2008. Following these activities extension officers trained in the project's train-the-trainer program established FFSs with farmer groups in 2009 (6 in villages in Fiji and 3 at villages in Samoa). The curriculum pursued in these FFS was very narrow and it was apparent at the mid-term review (August 2009) that extension officers would benefit from more FFS training. A master trainer has been engaged and will conduct intensive training activities (including curriculum development and reporting) with extension officers July-September 2010. FFS groups that have been developed by local extension officers will be integral to these training activities.

Studies to determine the role of major natural enemies in managing Brassica pests have yielded somewhat different results in the two countries. Regional surveys have confirmed that the DBM parasitoids Cotesia vestalis and Oomyzus sokolowskii are widely established in the major Brassica crop growing areas of Viti Levu, Fiji, and Upolu, Samoa. Exclusion studies have shown that natural enemies can have a significant impact on pest populations in both countries. However, while the impact of C. vestalis and O. sokolowskii on DBM populations have been quantified in Fiji, similar experiments have not yet quantified the impact of these parasitoids in Samoa; these studies are planned for July- August 2010. In Samoa significant progress has been made investigating the impact of the egg parasitoid Trichogramma chilonis on LCM. Regular field studies from November 2007 to February 2010 have shown that the parasitoid can have a very significant, but variable, impact on LCM eggs (marginal mortality rates vary from 0 to 0.474). Further, natural enemy exclusion studies have shown that neonate and second instar LCM larvae suffer significant mortality from predation. The challenge now is to understand the source of the variability of the impact of T. chilonis on LCM so that its considerable potential as a biological control agent can be exploited. This will enable the development of integrated pest management strategies that have the biological control of both major pests of Brassica crops (DBM and LCM) at their core.

Laboratory studies have confirmed that in Fiji DBM has developed resistance to several of the conventional insecticides currently used by farmers. Field populations collected from Sigatoka (upper, mid and lower valley), Fiji Agricultural College (FCA), Koronivia and Labasa (Vanua Levu) in 2009 all displayed resistance when compared to a standard laboratory strain of DBM. The three populations from the Sigatoka valley and the population from Labasa all exhibited resistance to deltamethrin and indoxacarb, while the population from FCA only displayed resistance to lufenuron. When the Sigatoka lower valley population was selected with either deltamethrin or indoxacarb in the laboratory, very high levels of resistance (relative to the susceptible Waite strain) could be selected within just three generations. These highly resistant populations demonstrated cross-resistance to deltamethrin or indoxacarb respectively but neither exhibited reduced susceptibility to either Bt (Dipel) or Prevathon. Laboratory selected resistant insects developed more quickly than individuals from both the Waite strain and the parent field strain but suffered greater mortality and produced smaller pupae, indicating that there may be a fitness cost to high levels of resistance.